tryon-tray 0.2.4.1

A package for virtual try-on services integration

Tryon-Tray

A Python package for virtual try-on services integration, supporting multiple VTON (Virtual Try-On)providers.

Installation

pip install tryon-tray

Usage

# Using Alphabake's virtual try-on API

from dotenv import load_dotenv
from tryon_tray.vton_api import VTON
load_dotenv()

result = VTON(
    model_image="inputs/person.jpg",
    garment_image="inputs/garment.jpeg",
    model_name="alphabake",
    auto_download=True,
    download_path="result_alphabake.png",
    show_polling_progress=True,
    # Alphabake specific parameters
    mode="balanced",  # Options: 'fast', 'balanced', or 'quality'
    garment_name="garment-example",
    tryon_name="tryon-example"
)

print(f"Result saved to: {result.get('local_paths')}")
print(f"Time taken: {result['timing']['time_taken']}")

Exploring Available Models

# Discover available models and their parameters
from tryon_tray import get_available_models, get_model_params, get_model_sample_config

# List all available models
all_models = get_available_models()
print("Available models:", all_models)

# List just the virtual try-on models
vton_models = get_available_models(category="vton")
print("VTON models:", vton_models)

# List video models
video_models = get_available_models(category="video")
print("Video models:", video_models)

# Get parameters for a specific model
alphabake_params = get_model_params("alphabake")
print("Alphabake parameters:", alphabake_params)

# Get a sample configuration for a model
sample_config = get_model_sample_config("klingai")
print("Sample KlingAI config:", sample_config)

# Use the sample config directly
from dotenv import load_dotenv
from tryon_tray import VTON
load_dotenv()

config = get_model_sample_config("fashnai")
result = VTON(model_name="fashnai", **config)

Features

• Multiple VTON service providers support
• Automatic image downloading
• Progress tracking
• Model discovery and parameter exploration

Configuration

Create a .env file with your API keys:

FASHNAI_API_KEY=your_fashnai_key
KLINGAI_API_KEY=your_klingai_key
REPLICATE_API_TOKEN=your_replicate_token
ALPHABAKE_API_KEY=your_alphabake_key

Sample Response

{
  "urls": ["https:/..."],  // Generated image URLs
  "local_paths": ["path/to/downloaded/image.jpg"],  // Downloaded file paths
  "timing": {
    "time_taken": datetime.timedelta  // Total processing time
  }
}

Parameters

• model_image: Path to the person/model image
• garment_image: Path to the garment image
• model_name: Service provider ("fashnai", "klingai", "replicate", "alphabake")
• auto_download: Automatically download generated images
• download_dir: Directory for downloaded images
• polling_interval: Time between status checks (seconds)`
• show_polling_progress: Show progressbar during generation
• category: Garment category ("tops", "dresses", etc.)
• mode: Generation mode ("quality" or "speed" for most APIs, "fast"/"balanced"/"quality" for alphabake)
• adjust_hands: Adjust hand positions in output
• restore_background: Preserve original image background

Response Format

• URLs! for generated images
• Local paths! to downloaded images
• *Timing! information (time taken for processing)

License

MIT License

Contributing

Contributions are wellcome! Please feel free to submit a Pull Request.